Kingston HyperX 1600MHz 12GB Triple Channel (KHX1600C9D3K6/12G)

INTRODUCTION

With RAM kits we've known for a long time that more RAM is better than less RAM. By that we mean if you have a choice between 3GB and 6GB or even 12GB get as much RAM as you can afford right out of the gate when building or upgrading a system. We've been testing i5 kits with 4 and 8GB and the 8GB kits give you a nice performance edge over 4GB kits. Now we are testing a Kingston HyperX 12GB kit so we are going to run the normal battery of tests on it then split the kit to a 6GB configuration and run some additional tests. So we'll have the same exact kit running in 6 and 12GB configurations in selected tests to show you that more ram is advantageous.

We believe that having a larger kit is so important that we can say it's more important to have more ram than it is to have faster ram. By that we mean if you have the choice of an uber expensive 6GB 2GHz kit or a similar priced 12GB kit that runs at a slower speed you are going to be better off buying the larger slower kit.

Larger kits provide more ram for system operations like Sleep, Wake, and loading large programs or tons of images for editing. Some people are thinking yea but the 3GB kit I have does the job. Yes it will do the job but you increase page file hits and the Page File resides on your HD which runs much slower than the RAM does. How much slower? Typical platter drives of modern design run in the neighborhood of 90MB/s under ideal conditions. RAM on the other hand runs in Gigabytes per second. Mid range kits running at 1600MHz can hit speeds of 18 20GB/s. That's a big speed difference. The larger RAM kits will allow your system to go to the Page File less and increase overall system performance. Here's the sad part, in synthetic benchmarks the smaller faster kit will come out ahead almost every time because the standard RAM testing utilities only use a small portion of the RAM kit to do testing. Which is why high end high speed kits get so much attention. They produce nice synthetic benchmark numbers. Real life performance the difference between a 1600MHz kit and a 2GHz kit is negligible. That's right real life performance between a 1600MHz kit and a 2000MHz kit is negligible meaning that you won't notice a performance difference between the two that the human eye can see. So if you have the choice between that 6GB 2GHz kit you have been lusting after or a 12GB kit go for the 12GB kit. It would be hard to make a better choice than the Kingston HyperX 1600MHz 12GB kit we are looking at today.

Specifications

• Capacity 6 x 2GB (12GB Total)
• Speed DDR3 1600 (PC3 12800)
• XMP Profile
• Model Number KHX1600C9D3K6
• 240-Pin DDR3 SDRAM
• Cas Latency 9
• Voltage 1.65V DRAM 1.4v QPI
• 240-Pin DDR3 SDRAM
• Hexa-Channel Kit

We can see those wheels turning in your heads already. Why not a 3 x 4GB module kit? The answer to that is simply price. A 3 x 4GB kit of HyperX with the exact same specifications currently runs $1299 at reputable E-Tailers, the 6 x 2GB kit runs $359 so it's more affordable for most end users. We can see situations where the 3 x 4GB kit would be advantageous like dedicated Video Editing but for most of us the $1299 is just to much for a 12GB kit.

Now once again we can see the wheels working. Why a kit with 9-9-9-27 timings and not two kits of 8-8-8-24? Voltages on the chipset get higher as the ram gets faster and the timings get tighter.This kit that runs at 9-9-9-27 will take 1.65v DRAM voltage and 1.4v QPI while a kit of 8-8-8-24 (6 GB) will take the same 1.65v DRAM but take more QPI voltage (in most cases). Ram kits are already pushing the safe limits of QPI voltage. How do we know that kits are pushing the safe limits of QPI voltage, tons of research, the maximum voltage according to Intel for QPI is 1.35v. You can read the Intel Whitepaper on the i7 965 Here pay attention to Page 22, or here's an excerpt we pulled out for you.

"Voltage for the digital portion of the integrated memory controller, QPI link and Shared Cache with respect to VSS. 1.35v (Max)."

NOTES

• 1. For functional operation, all processor electrical, signal quality, mechanical and thermal specifications must be satisfied.
• 2. Excessive overshoot or undershoot on any signal will likely result in permanent damage to the processor.
• 3. VTTA and VTTD should be derived from the same VR.

So Intel is saying that 1.35v is the maximum safe voltage for QPI and kits often exceed that Intel specification. We haven't seen any boards or CPU's die from a slight voltage bump over 1.35v but Intel built it and they are saying 1.35v so rather than landslide us with E-Mail about "OMG my kit runs over 1.35v" please direct those E-Mails to Intel who is a lot more qualified to answer your panic stricken E-Mails than we are. We don't mind running above the 1.35v ourselves it hasn't killed any CPU's yet but that's what Intel says.

Now back to our Kingston HyperX kit. This kit runs at 1.65v DRAM and 1.4v QPI and we have confidence that it's safe to run at that voltage. Now take two kits of 6GB ram running at 1.65v Dram and 1.5v QPI designed to run as a standalone kit with 6GB and put in two kits and you might have to run the QPI even higher. A single kit of 12GB designed to run as a 12GB kit is less likely to make you push QPI and will (in the long run) be safer and more likely to function plug and play like our kit did.

That being said we have kits that run at and above 2000MHz that demand 1.6v and even 1.65v QPI and it hasn't killed our CPU's (YET). These things tend to work long term so eventually it might damage one but by that time we will have a newer faster shinier CPU. You might want to consider longevity on the i7 platform though it's a little pricey.